With the shortcomings of traditional algorithm in video surveillance on low accuracy, poor robustness and unable achieved real-time tracking for multi-targets, this paper presents a Multi-target tracking algorithm, DeepSort, on the base of deep neural network to achieve the end-to-end surveillance video multi-personal target real-time detection and tracking. The high accuracy of target detection by YOLO algorithm provides DeepSort with weaker dependence on detection results, lower interference of occlusion and illumination and improved tracking robustness. Moreover, due to the high redundancy of the surveillance video itself, the difference filter is used to screen the video frames with no foreground targets and small changes, so as to reduce the detection cost and improve the detection and tracking speed. The experimental evaluation of the video surveillance dataset NPLR, the average MOTA of this algorithm is 68.7, the highest value is 86.8; the average speed is 81.6Hz, the highest value is 140Hz. It shows that the end-to-end algorithm is feasible and effective.
We present a novel method to segment an object from multiview images using level set method. Our approach takes advantage of the unique property of level set method in the flexibility of objective energy function design and the adaptability to cut boundary with arbitrary topology. We introduce an iterating optimized 3D level set framework for view coherent segmentation and propose three forces in this framework to drive the convergence of level set to the ideal boundary. In between, the point cloud term and the edge term are designed to give an as-good-as-possible boundary indicator for the level set function, while the local color discriminative classifier is iteratively updated with the multiview silhouette and the 3D point cloud to drive the deformation of the zero level set. Extensive experimental results demonstrate that our approach can produce much more accurate edge localization and more coherent segmentation result across views, compared with the state-of-the-art methods, even for the case of very challenging foreground topologies and ambiguous foreground-background color distribution.
Strap-down inertial navigation system (SINS) is widely used in military field, to facilitate the study of SINS algorithms and various coupled navigation algorithms, a simulation system of SINS is designed. Based on modular design, with good portability and expansibility, the system consists of four independent modules: analysis module of motion state, trajectory simulator, IMU simulation module and SINS calculation module. With graphical interface, the system can control every motion state of the trajectory, which is convenient to generate various trajectories efficiently. Using rotation vector attitude algorithm to process simulation data, experiment results show that the attitude, velocity and position error is consistent with the theoretical value, which verifies the rationality of the simulation model and the availability of the simulation system.
Surface physical damage detection is an important part of the shaft parts quality inspection and the traditional detecting methods are mostly human eye identification which has many disadvantages such as low efficiency, bad reliability. In order to improve the automation level of the quality detection of shaft parts and establish its relevant industry quality standard, a machine vision inspection system connected with MCU was designed to realize the surface detection of shaft parts. The system adopt the monochrome line-scan digital camera and use the dark-field and forward illumination technology to acquire images with high contrast; the images were segmented to Bi-value images through maximum between-cluster variance method after image filtering and image enhancing algorithms; then the mainly contours were extracted based on the evaluation criterion of the aspect ratio and the area; then calculate the coordinates of the centre of gravity of defects area, namely locating point coordinates; At last, location of the defects area were marked by the coding pen communicated with MCU. Experiment show that no defect was omitted and false alarm error rate was lower than 5%, which showed that the designed system met the demand of shaft part on-line real-time detection.
The packing presswork is an important factor of industrial product, especially for the luxury commodities such as cigarettes. In order to ensure the packing presswork to be qualified, the products should be inspected and unqualified one be picked out piece by piece with the vision-based inspection method, which has such advantages as no-touch inspection, high efficiency and automation. Vision-based inspection of packing presswork mainly consists of steps as image acquisition, image registration and defect inspection. The registration between inspected image and reference image is the foundation and premise of visual inspection. In order to realize rapid, reliable and accurate image registration, a registration method based on virtual orientation points is put forward. The precision of registration between inspected image and reference image can reach to sub pixels. Since defect is without fixed position, shape, size and color, three measures are taken to improve the inspection effect. Firstly, the concept of threshold template image is put forward to resolve the problem of variable threshold of intensity difference. Secondly, the color difference is calculated by comparing each pixel with the adjacent pixels of its correspondence on reference image to avoid false defect resulted from color registration error. Thirdly, the strategy of image pyramid is applied in the inspection algorithm to enhance the inspection efficiency. Experiments show that the related algorithm is effective to defect inspection and it takes 27.4 ms on average to inspect a piece of cigarette packing presswork.
The airborne LiDAR system, which usually integrated with optical camera, is an efficient way of acquiring 3D
geographic information and enjoys widely application in building DSM. However, when the airborne LiDAR is used in
urban area, where there are a large amount of tall buildings, the characteristic points of buildings are seldom measured
and the measured points are frequently too sparse to create precise building models. In this paper, an approach to DSM
refining DSM in urban area with fusion of airborne LiDAR point cloud data and optical imagery is put forward. Firstly,
the geometric relationship between the airborne LiDAR point and the correspondent pixel on the image synchronously
taken by optical camera is analyzed. The relative position and attitude parameters between the laser rangefinder and the
camera are determined in the process of alignment and calibration. Secondly, the building roof edges on the optical
image are extracted by edge detection. By tracing the building roof edges, the contours of building roofs in vector format
are acquired and the characteristic points of buildings are further extracted. Thirdly, all the LiDAR measured points on
the roof of specific building are separated from the point cloud data by judging the geometric relation between LiDAR
measured points and the building outline, which is represented by a polygon, according to their plane coordinates.
Finally, the DSM refinement for buildings can be implemented. All pixels representing the building roof are given
heights as same as that of nearer LiDAR point inside the polygon. Ortho-photo map and virtual building models of urban
area with higher quality can be reached with the refined DSM and optical images.
KEYWORDS: Calibration, LIDAR, Clouds, Airborne laser technology, 3D metrology, Mathematical modeling, Data modeling, Imaging systems, System integration, Global Positioning System
Airborne Lidar measurement technology, as an efficient way of acquiring three-dimensional geographic information,
plays an important role in building DSM and DEM rapidly. Because the airborne Lidar measurement system usually
integrates multiple devices including GPS receiver, INS, laser rangefinder and CCD camera, the relative geometric
position and attitude relationships among these devices must be accurately measured in order to get the points with high
precision and thereby satisfy the accuracy requirements of produced DSM and DEM. It is proved that the misalignment
of airborne Lidar system, which is represented by angle deviations of yaw, pitch and roll, is the most significant source
of systematic error in airborne Lidar measurement. In this paper, the effect of pitch angle error on the 3D coordinates of
measured point is firstly analyzed. On this basis, a calibration method of the pitch angle deviation for airborne Lidar
system by using the geometric characteristics of spire houses is put forward. The proposed pitch angle deviation
calibration method consists of four key steps: (1) Initial pitch angle calculation. In the light of the offset distance
between the ridge lines of the same house acquired by airborne Lidar system flying in opposite directions, an initial pitch
angle deviation can be calculated. After separating the effect of pitch angle deviation, the rectified laser point cloud data
are obtained. (2) Roof plane equation determination. The plane equations of both roof slopes are determined by fitting
algorithms with the 3D coordinates of points located in the same spire roof. (3) Distance standard error calculation. The
distance of each point to the roof plane is computed and applied to the calculation of distance standard error. (4) Final
pitch angle deviation calculation. Taking the distance standard error as the overlapping criterion, the pitch angle
deviation correction is iteratively calculated according to the aforesaid procedure until the distance standard error is less
than a given value. The final pitch angle deviation is the sum of all the pitch angle deviation corrections. Experiments
show that the proposed calibration method is correct and effective.
This paper introduces the theories and methods for image fusion in remote sensing. Based on the IHS transform and wavelet packet transform, a scheme of remote-sensing images fusion aiming at protecting image's spectral characteristics is made in this paper. This scheme makes best use of the information in remote-sensing images to be fused and prevents the loss of image information. Through the experiment using SAR image and TM image, all the spectral characteristics, the textural feature and the spatial quality of image have been improved.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.